UNIVERSIDADE FEDERAL DE SANTA <
PÓS-GRADUAÇÃO EM INGLÊS E LITERATURA, CORRESPONDENTE
THE (UN)CONSCIOUSNESS OF SECOND LANGUAGE ACQUISITION
de
PAULA ALEXANDRA FATÜR SANTOS
Dissertação submetida à Universidade Federal de Santa Catarina para obtenção do grau de MESTRE EM LETRAS
FLORIANÓPOLIS Setembro de 1994
MESTRE EM LETRAS
Qpçâo Inglês e Literatura Correspondente
Prof* Dr8 Carmen Rosa Caldas-Coulthard COORDENADORA
BANCA EXAMINADORA:
Prof Dr8 Bárbara O. Baptista
Prof Dr. José Carlos Paes de Almeida Filho
To my father, who said 'go for it', arid to Mauro, who came along.
Having grown in a home where both parents are bilinguals lias sparked in me a spontaneous interest in language. Having studied a second language in an educational setting has raised in me a high awareness of language. In a sense, this survey represents the realization of gradual, evolving exposure to second language: from home to school to research. Theories and claims about second language acquisition (SLA) have rarely been contrasted, despite the crucial role that they play in the development of a second language, either in a natural or educational environment. It is only recently that classifications of SLA theories have appeared in the literature, showing an attempt by researchers to find similarities and differences among existing theories. This study reviews the process of second language acquisition in SLA theories and discusses the notion of consciousness in the theories. A few comments are in order here.
First, only theories which deal with the cognitive aspect of learning are included. Second, the chapters which discuss these theories can be read in any sequence the reader may wish for they are modular in nature. And as they are self-contained, I decided to encapsulate the bibliographical references within each chapter. Third, the order in which the theories are presented resembles the order in which they were written. This somehow reflects the growth of the researcher in terms of writing ability, development of critical view, and power to analyse the theories. Fourth, it was felt that the abundance of technical terms might interfere with comprehension. Thus, for purposes of enhanced readability, a glossary for technical terms is provided at the end of tins study.
I am very grateful to my advisor, Hilário Bolin, for providing an enriching environment of academic excellence that was crucial to this work.. I am also thankful to Prof. Michel Paradis, who kindly listened to my questions and answered them. Finally, I would like to make mine the words of a successful neurosurgeon from New York University Medical Center, Dr. Fred Epstein: 'We need the courage to pursue new concepts and ideas, otherwise the fear of being wrong keeps us from ever being right’.
THE (UNCONSCIOUSNESS OF SECOND LANGUAGE ACQUISITION PAULA FATUR SANTOS
UNIVERSIDADE FEDERAL DE SANTA CATARINA 1994
Supervising Professor: Dr Hilário Inácio Bohn ABSTRACT
This theoretical study aims at presenting what some important psycholinguistic theories of SLA say about the process of SLA in general, and about the conscious and unconscious aspects of this process in particular. It also compares the psycholinguistic view found in the SLA field with a cognitive psychological view, Ausubel's Assimilation theory. Similarly, it compares the psycholinguistic view with a neurolinguistic view, Paradis’. Such multi-disciplinary comparisons were provided because it is assumed that the SLA process is highly complex, involving multiple aspects and having to draw from multiple areas in order to be fully explained. The study results in a classification of psycholinguistic theories of SLA according to their source-influence areas. Thus, theories were classified into Linguistics-based and Psychologically-based theories. The classification brings to light the two trends that permeate recent psycholinguistic research in SLA one that argues for a conscious type of learning, the psychological trend; and one that argues for unconscious acquisition, which has its roots in Chomsky, the linguistic trend.
PAULA FATUR SANTOS
UNIVERSIDADE FEDERAL DE SANTA CATARINA 1994
Professor Orientador: Dr. Hilário Inácio Bohn RESUMO
Este estudo teórico analisa o processo de aquisição de segunda língua, em geral, e os aspectos conscientes e inconscientes deste processo, em particular, nas mais importantes teorias psicolingüísticas de aquisição de segunda língua Além disso, este mesmo processo é analisado em uma teoria psicológico-cogniiiva, a teoria da Assimilação de Ausubel, e em uma teoria neurolingüística, a de Par adis. A busca de teorias multidisciplinares deve-se ao fato de que o processo de aquisição de segunda língua è per se altamente complexo, envolvendo múltiplos aspectos e, conseqüentemente, múltiplas disciplinas. A análise do processo nas teorias culmina numa classificação das mesmas. A percepção das áreas subjacentes às teorias foi o critério encontrado para classificá-las. Assim, elas foram classificadas em teorias que emanam da Psicologia, tendência esta que privilegia o aprendizado consciente, baseada na Psicologia Cognitiva, e que emanam da Lingüística, tendência que privíligia a aquisição inconsciente e que tem suas raízes
vli TABLE OF CONTENTS
Preface... ... ... iv
Abstract... ... v
Resumo... vi
Table of Contents... vii
List of Figures...x
Chapter 1 INTRODUCTION... ... . 1
References ... ...9
Chapter 2 DRAWING FROM PSYCHOLOGY... ...10
The nature of unconscious processes and contents... 10
Chomsky's view of the unconscious... 11
Unconscious processes and knowledge...12
Ausubel's Assimilation Theory... 13
References ... ...17
Chapter 3 THE MONITOR MODEL... ...18
The model... ... 18
The cognitive ability...19
The non-interface hypothesis... 21
The conscious/subconscious issue... 22
Empirical evidence...23
References ...26
Chapter 4 THE TASK-BASED APPROACH... ...27
The system... ...2J The process... .... 29
Language awareness... ... 29
The pedagogy of the task-based approach... ...30
The cognitive counterpart... ... ..31
The cognitive underpinnings of the task: problem-solving...32
References ... ...:... 34
Chapter 5 THE MULTIDIMENSIONAL MODEL... 35
The developmental dimension... ... 36
The variational dimension... . 37
The. theoretical constructs of Pienemann's model... 40
Concerning the acquisitional view... 40
Concerning the processing strategies... 40
Seliger’s view of the acquisition process... ... 46
Linguistic knowledge... 46
Metalinguistic knowledge... ... 47
Verbal reports in Seliger’s view...48
Verbal reports and other areas... ...49
References ... ... ...52
Chapter 7 SHARWOOD SMITH... ... . 53
Early View... ... 53
Linguistics and language pedagogy... 53
The acquisition process... 54
Metalinguistic awareness... 56
Consciousness-raising... ... 57
Input enhancement... ... 59
Rjecent View: Language processing...60
References ... ... ... 68
Chapter 8 INFORMATION-PROCESSING THEORY... 69
Hie psychological underpinnings:... ... 69
Shiffiin & Schneider’s theory... ... 69
Rumelhart & Norman’s 'Restructuring'...73
The psychological view applied to language learning...74
Practice or rote learning ?... ... 80
The consciousness issue... 81
References ... ... ... 85
Chapter 9 CLASSIFYING SLA THEORIES... ... . 86
The controversy... ... ... .87
The classifications... 91
The proposed classification... ... 100
References ... . 103
Chapter 10 DRAWING FROM NEUROLINGUISTICS... . 105
The organization of two languages in the brain... 106
Neurofunctional modularity of the systems... 108
Memory systems involved in the acquisition of language...110
Implicit competence...112
Explicit knowledge... 114
IX
Evidence for dissociation between declarative and procedural
memory...116
Maturational constraints on memory systems... 119
Language aspects belonging to procedural and declarative memory... ... 120
Metalinguistic knowledge...122
The limbic system... ... 125
References ... 129
Chapter 11 CONCLUDING REMARKS... ...130
Pedagogical implications... 133
GLOSSARY...135
Fig. 1 Factors involved in the learning process,
highlighting the ones covered in this dissertation... . 2
Chapter 6 Fig. 1 How verbsd reports are used in Cognitive Science and in Language Processing research...50
Chapter 7 Fig. 1 Sharwood Smith's view of the interlanguage system...56
Fig. 2A Sharwood Smith's bilingual modular mental processing model Central Processor... ...64
Fig. 2B Sharwood Smith's bilingual modular mental processing model Linguistic Module... ... ...65
Chapter 9 Fig. 1 McLaughlin's debates...91
Fig.2 Larsen-Freeman's classification... 92
Fig. 3 Bialystok's classification... ...93
Fig. 4 Snow's classification... ... ... 95
Fig. 5 The proposed classification... 100
Chapter 10 Fig. 1 Types of memory, according to Paradis (1993). ... I ll Fig.2 Characteristics of procedural competence and declarative knowledge, based on Paradis (1993)... 115
CHAPTER 1
INTRODUCTION
With the explosion of Cogni&vism, marked by Piaget in Psychology and by Chomsky in Linguistics, the behavioral sciences entered a new era. Studies shifted from behavior as the object of studies to behavior as the means to understand the workings of the mind. Thus, the emphasis was placed on the mind, its processes, what underlying reasons cause a certain behavior.
Research in Second Language Acquisition (SLA) followed the above tendency, moving from research on teaching to research on the learning process. That move also marked the emergence of SLA as an independent discipline, when numerous issues related to the acquisition of a second language (SL) started to be studied. As a consequence, the field started to draw from the most diverse areas of knowledge -- Psychology, Linguistics, Sociology, Psycholinguistics, Sociolinguistics, Neurolinguistics (cf. Larsen-Freeman & Long 1991:5) - and is expected to collaborate with these areas as well.
This muKidisciplinarity was especially beneficial to the study of the process of second language acquisition. This process involves a diversity of variables, making it a very complex and controversial issue. The number of SLA theories that have been proposed in the literature is evidence of this. Larsen-Freeman & Long (ibid:227) suggest that there are over 40 theories of SLA Each of these theories tries to explain acquisition from a specific viewpoint, be it the social, the affective, the cognitive, or some other. And even within the scope of a specific viewpoint, say, cognitive theories, there is controversy. For one tiling, theories that follow a Chomskyan tradition often claim that the process is largely unconscious, a matter of triggering what is innately programmed. On the other hand, Hie cognitive-psychological tradition attributes much
consciousness to Hie process, all of which comes from experience. In any case, perhaps precisely because of its complexity and controversial nature, the SLA. process is so amazingly interesting to researchers.
The complexity of the acquisition process can be better visualized when we consider the factors involved in it, such as the neurological, psychological, and social factors . All factors are sufficiently important in SLA to engage our most serious considerations, as they influence the learning process concomitantly and interact in various ways. Nevertheless, either set can be examined while the others are held constant. This applies to this present study of cognitive processes — it looks at just a fraction of the whole learning process, but a fraction which seems to be more immediate, less tangential, more crucial in its impact on SLA than are affective and social factors. Figure 1 spells out the main factors involved in the SLA process:
Fig.l Factors involved in the learning process, highlighting the ones covered m this dissertation.
3 This theoretical study aims at presenting what some important psycholinguists theories of SLA. say about the process of SLA in general, and about the conscious and unconscious aspects of this process in particular. It also compares the views found in the SLA theories with a cognitive psychological view, namely, Ausubel’s Assimilation theory. The comparison is due to a belief that a second language acquisition theory should draw from theories of learning and not only from theories of language since in many aspects cognitive development and language development overlap. Similarly, this study compares the psycholinguistic view with Paradis' neurolinguistic view, in order to provide a realistic neurolegically feasible picture of the acquisition process. The study results in a classification of psycholinguistic theories of SLA according to the areas which constitute their source of influence, the criterion which was found to be the most relevant for the clarification of the psycholinguistic picture of SLA.
Cook (1993) describes the three aspects that a (psycholinguistic) theory or model of SLA should cover.
1. the description of the competence system, or mental grammar, in the Chomskyan sense, or knowledge about the language, as it has been called more recently;
2. the description of processes used to build (acquire) the competence system, or developmental processing, as called by Sharwood Smith (1991); and
3 the description of procedures to access the competence system, or the ability to use language, or, following Sharwood Smith's terminology, on-line knowledge processing.
In the following lines, I use Cook's description to determine the scope of tins dissertation.
Concerning aspect number 1, when I say that a theory includes a description of the competence system in whatever form it may be, I do not mean only a description of the linguistic system as such, of its phonology, mofphosyntax, and semantic constraints on the lexicon (the generative), or even a description of the network of information which constitutes language competence (the connectionist), rather, I also mean a description of the nature of the system, what Hie competence system consists of. Note that, when describing the nature of the system, I am not leaving aside the territory of metaphors and abstractions which characterize the descriptions of the competence system as such, like the generative and connectionist descriptions, but I am somehow going beyond the theoretical territory and going towards a realistic picture of how that competence system is realized in the mind, conceived as a real entity, made feasible. It must be clear that this realization (in the sense used above) is not free of metaphors and abstractions, for not only psychology but also neurology has to rely on metaphorical models and abstractions in order to explain how the mind works.
In some theories, aspect number 1 is inevitably linked to number 2, that is, the description of the acquisition processes and that of the representation of the linguistic system are inseparable (for instance, Sharwood Smith 1991). In others, aspect number 1 is implicit, that is, if the processes described in the acquisition of the linguistic system are said to be similar to Chomsky's description of LI processes, we expect that this linguistic system refers to the generative /Universal Grammar model. This is the case in Krashen’s and Prabhu's theories. Thus, aspect number 1 needs to be considered in order to show whether a theory stems from linguistics or psychology.
Aspect number 3, language processing, deals mainly with language use, a processing phenomenon that comprises the linguistic tasks of production and reception. The use of unconscious processes in these tasks is undeniable and
5
has been more aid more described by psycholinguists (Jackendoff 1987; Sharwood Smith 1991). Cognitive psychologists Meichenbaum and Gilmore (1984:276) compare such processes to physiological processes:
In the same way that we engage in a variety of physiological processes (breathing, stomach contractions, and so foith) without awareness, a similar analogy can be drawn to how we process information.
Misunderstandings involving aspect number 3 happen sporadically in theories that claim to be learning theories, when in fact they address language processing. This seems to be the case with McLaughlin (1987) in the first version of his theory, where he explained basically access to knowledge, but claimed to be presenting a cognitive theory of SLA. Because of criticisms of this kind (Cook 1993:267), his more recent version (1990) contains the notion of restructuring, a notion devised to explain knowledge acquisition. But in general, theories of language processing are well defined and self-contained so that at first it would seem possible to leave them out of a study that does not specifically address language processing, if it were not for one reason. The acquisition of language competence is only shown in performance. Thus, as is the case ip. this study, sometimes it is impossible to talk of language acquisition without talking of performance mechanisms.
While aspect number 1 deals with the description of the competence system, and aspect number 3 deals with procedures to access this system, in this dissertation I am concerned with aspect number 2, acquisition of the competence system.
Some clarifications about terminology seem in order. Hie first concerns the theories dealt with in this dissertation. I chose to call theories of SLA that dead with the cognitive aspect of the learning process psycholinguistic theories of SLA My choice is principled, and it is explained in chapter 8, where I discuss
the recent polemics that involve the word psycholinguistics. The second concerns the use of the expression SLA. The range of SLA as an independent discipline has broadened significantly recently, and presently the term. SLA encompasses the study of both a second and foreign language environments, irrespective of the differences between them. The same is true for the key terms acquisition and learning, the third terminological problem to be clarified. Acquisition has become a superordinate term for both of them, although originally acquisition refers to natural settings and learning to classroom settings. In this study, Hie terms are used interchangeably, irrespective of the differences between them, except where it is indicated that a specific meaning is ascribed to them.
Theories are first explained as a whole — however briefly, the functioning of theories is explained — with special emphasis on the process of SLA. Then, the theoretical constructs of the theories are analyzed. These constructs provide tools for a discussion of the conscious and unconscious aspects of each theory, and they reveal the areas which constitute the source of influence of the different theories. From the areas, a natural classification emerges.
An outline of the dissertation follows:
Chapter 2, Drawing from Psychology, reviews some important psychological issues that are useful to the present debase of conscious versus unconscious processes in SLA, drawing from the area that is generally in charge of explaining them: psychology. Ausubel’s Assimilation Theory is also presented in this chapter in order to introduce the reader to a sound cognitive psychological theory of knowledge acquisition and acquaint the reader with his unique explanation of how knowledge is organized so as to form a cognitive structure.
7 Chapter 3 presents Krashen's Monitor Model. It discusses the distinction between acquisition and learning processes. It also evaluates how the notions of conscious/unconscious serve as the constructs of this theoretical model.
Chapter 4 deals with Prabhu's Task-oriented Approach. Actually, this is a pedagogical proposal rather than a SLA theory or model, but it was included in this study for the interesting tenets that underlie the approach and lead to the pedagogical procedures proposed, the tasks. In this chapter, thus, the tenets of his approach are discussed, as well as the cognitive value of the task for the acquisition process.
Chapter 5 introduces an experimental model — Pienemann’s Multidimensional Model. Initially, I consider his two dimensions of the learning process: the developmental and the variational. Then, I establish the theoretical construct of the model, namely, sentence processing.
Seliger’s view of the language acquisition process is presented in chapter 6. First, I present the two kinds of knowledge, linguistic and metalinguistic, that he claims are involved in SLA Second, I discuss his particular view of verbal reports and their validity for explaining certain complex issues.
Chapter 7 examines Sharwood Smith’s view. First, the Pedagogical Grammar Hypothesis which belongs to the early phase of his work is analyzed. The analysis thereon will concentrate on the more recent phase, which is influenced byFodor’s modularity thesis.
In chapter 8, McLaughlin's Information-Processing Model, the cognitive exemplar of the learning of a second language, is examined. First, I trace Cognitive Psychology as its source of influence. Second, I discuss his emphasis of language use over acquisition. Third, I discuss later incorporations to the model, such as the notion of restructuring, which were inserted in order to account for linguistic idiosyncrasies.
Chapter 9 has a speculative nature. It reaches the point of classifying the theories analyzed in terms of their source-infiuence areas in order to clear up the picture of psycholinguistic research in SLA. In this way, it is somehow conclusive as far as the theories studied are concerned.
Chapter 10, Drawing from Neurolinguistics, incorporates the substantial contemporary contributions of a neuroscience into the discussion of the SLA process. It presents a realistic picture of how linguistic information is organized in Hie brain and the mechanisms used to acquire and access it.
In the concluding chapter, I pull together the insights gained from the study of SLA theories and the insights provided by neurolinguistics, thus achieving a state of the art on the issue of psycholinguistic processes used in SLA Pedagogical implications of the proposed view are also considered.
Although I have attempted to describe the processes found in the various theories/models extensively and profoundly, this was not always possible due to the obscurity and subjectivity that permeate the theories. The subjectivity lies, among other things, in the lack of precision of terms, which are widely used but whose real meaning is never made clear to the reader. Thus, we are obliged to infer without being sure how much we can infer from what is said. The obscurity seems to be a reflex of the present state of research about the *black box' ~ it seems that authors are just feeling their way within the SLA field: hypothesizing, speculating, theorizing. While this happens, other fields (such as die Cognitive Sciences) start to take over research matters that previously belonged to SLA Or is it SLA that is moving towards these fields? In any case, a chance must be given for SLA research to show how much it has come to know about the ’black box’ in these twenty years of research as an independent discipline.
9
Cook, V. 1993. Linguistics and Second Language Acquisition. London: MacMillan Press.
Jackendoff, R. 1987. Consciousness and the Computational Mind. New York: Academic Press.
Larsen-Freeman, D. & M. Long. 1991. An Introduction to Second Language Acquisition Research. New York: Longman.
McLaughlin, B. 1987. Theories o f Second Language Learning. New York: Edward Arnold.
McLaughlin, B. 1990. Restructuring. Applied Linguistics 11(2): 113-28.
Meichenbaum, D. & J .B. Gilmore. 1984. The nature of unconscious processes: a cognitive-behavioral perspective. In: Bowers, K.S. & D. Meichenbaum (eds.) The Unconscious Reconsidered. New York: Wiley.
Sliarwood Smith., M. 1991. Language models and bilingual processing. In E. Bialystok (ed.) Language Processing in Bilingual Children. Cambridge: CUP.
DRAWING FROM PSYCHOLOGY
The value o f consciousness in human life is something which extends across philosophy to psychology and to linguistics.'
Vivian Cook (1993:68)
In this chapter, I discuss some important concepts to this dissertation borrowed from psychology. I first discuss the nature of unconscious processes and contents, as proposed by Reber in Experimental Psychology. I also take on Chomsky's view of the unconscious, although he is traditionally considered a linguist. Then, I present the concepts of unconscious processes and unconscious knowledge as they should be understood in this dissertation. Finally, I briefly explain Hie most typical exemplar of a cognitive psychological theory, namely,. Ausubel's Assimilation Theory.
The nature of unconscious processes and contents
According to experimental psychologist Reber (1989), devoted to the study of implicit learning and unconscious cognitive processes, consciousness is a mental state of late origin 'The proper stance is to assume that unconscious mental states are the foundations upon which emerging conscious operations are laid' (ibid:230). As to the nature of unconscious cognition, he assumes that we have a primitive unconscious that we share 'with all corticated species. And we have a sophisticated unconscious which has a basic difference from the former - - its processes, which all depend on a rich, abstract knowledge base which controls perception, affective choice and decision-making independently of consciousness. The primitive unconscious is active and fundamental for survival. Three points are important for the comprehension of the functioning of the primitive unconscious:
11
1. Hie pickup of information takes place in the absence of awareness of what is being picked up (implicit learning).
2. Much of what is acquired can be explained/ articulated/verbalized consciously, but the amount that is stored is much greater than the amount that can be explained.
3. The function of the primitive unconscious processes (which are automatic) is ’to pick up critical knowledge about categories and about co-variations of aspects of categories’ (ibid:231). Such knowledge is not related to meaning, affect, or interpretation ~ these belong to the sophisticated unconscious.
The sophisticated unconscious depends on previously acquired knowledge, whereas the primitive has the function of acquiring such knowledge. There are other differences between them. The sophisticated unconscious is generally available to consciousness, that is, there is awareness of the knowledge base itself. The scope of this awareness, however, is greater than the scope of the covert base. In other words, the possibility of information of the knowledge base becoming conscious is greater than of it remaining unconscious. These bases work at a symbolic level. They all involve semantic and affective properties. It is similar to what is generally known as cognitive structure.
Chomsky's view of the unconscious
Chomsky (1980 cited in Bowers & Meichenbaum 1984.156) considers unconscious processes those which are inaccessible in principle, that is, it is possible to ’acquire knowledge of their presence and activity by inference, but not by means of direct, immediate introspective awareness’. Unconscious knowledge (or tacit, innate) is in the mind, it is employed in the service of ongoing cognitive processing, but is not accessible to introspection. It is active, but not reachable. Besides, unconscious linguistic knowledge is basically the same for every speaker-hearer (universal).
If we compare Rebel’s definition and Chomsky's, we will see that in both senses of the word, the psychological and the linguistic, the unconscious is active rather than dormant; and in both cases, unconscious mental processes operate on different principles than conscious ones. But both definitions conform to the general principle in psychology (Bowers & Meichenbaum 1984:156) that there is an unconscious level of mental activity which influences behavior in one way or another .
Unconscious processes and knowledge
In order to provide the reader with a definition of the unconscious congruent with the notion used in tins study, I will take the meaning that is most often implied in learning theories as well as in language theories. When I speak of the unconscious in such theories, I am referring either to the unconscious process o f learning or to the product of this process, unconscious knowledge. Mien the process is under consideration, then this is implicit learning, a type of learning of complex skills, such as a language, held outside awareness. When we refer to the product of tins process, it is tacit knowledge that we are referring to, knowledge that is internalized but not verbalizable. It is possible to relate such notions with those posed by Reber as to the nature of unconscious processes: implicit learning involves the use of the primitive unconscious. The result of implicit learning is tacit knowledge, whose probability of being unavailable to verbalization is high.
Actually there is no clear border line as to which cognitive processes used in acquisition are unconscious and which are conscious. Claims are often the result of speculation of researchers. What we can say is that some processes have a higher probability of being used unconsciously, although it is safe to say that tins is polemical. Here I list some of them, frequently cited in psychological and linguistic literature (Ausubel 1978; Snow [in press], McLaughlin
1978:321-13 3): induction, deduction, inferencing, hypothesis-fonnulation and testing, transfer, acquisition heuristics (as posed by McLaughlin:ibid) such as overgeneralization, avoidance, imitation, simplification, Slobin's operating principles (see glossary). Problem-solving can be considered an unconscious process when it is a means to achieve a mag or goal. When it is a goal per se, it is a conscious strategy. But in that case, the step-by-step procedures that lead to the solution of the problem are unconscious, since the focus of attention on the goal to be achieved turns the attention away from the procedures.
In this study, it is presupposed that both conscious and unconscious processes are involved in SLA, but it is asked what role each of them plays in the SLA process and how they are characterized in SLA theories.
Having discussed some concepts relevant to the rest of this dissertation, I will now describe a cognitive psychological view of the acquisition of knowledge. The outstanding name here is Ausubel, who has the most complete, unparalleled theory, about the acquisition of knowledge, the Assimilation Theory.
AusufoeFs Assimilation Theoiy
David Ausubel has had a significant impact on our present understanding of the process of learning, particularly learning in an educational setting. His book Educational Psychology - A Cognitive View is a classic in the field of educational psychology. The Assimilation Theory, as labeled by Ausubel and his associates, contends that the learner's present knowledge plays a critical role in the process of future learning. His theory of learning has a more cognitive focus than an affective-social one, and so the contribution of his thought may be fully appreciated in the context of discussion on cognitive variables in learning.
One of the model’s greatest strength qua theory is that all knowledge is hierarchically organized in the learner's mind and that this prior structured knowledge is the most important factor at the time of learning, determining the
learner's capacity for acquiring new concepts or information. Ausubel (1978:163) puts it this way:
If I had to reduce all of educational psychology to just one principle, I would say this: The most important single factor influencing learning is what the learner already knows. Ascertain this and teach him accordingly.
Prior knowledge in this case is characterized as the quantity and organization of the learner's present knowledge which is made up of information, concepts, ideas, facts and propositions. This body of knowledge constitutes the learner's cognitive structure. Ausubel maintains that the most inclusive ideas are set at the top of this structure, thus maintaining a hyponymical organization.
Based on his observations of the learning process, Ausubel assumes that learning takes place in the student's mind through a process of meaningfully relating new concepts to old concepts in cognitive structure. For the present, perhaps, it is necessary to distinguish the properties of cognitive structure that influence learning.
First, the learning of new information is a function of the existence of relevant anchoring ideas. These anchoring ideas occupy a more general, inclusive position in the structure of knowledge and serve as cognitive hooks or pegs on which new ideas can be hung.
A second major property in learning is the discriminability of new concepts from old anchoring ideas in cognitive structure. The assumption here is that existing knowledge is dominant and new knowledge is subservient whenever they are similar in nature. In such a case, old knowledge takes over the cognitive field and superimposes itself on similar new knowledge.
A last factor affecting learning is the stability and clarity of the established anchoring ideas. A stable and clear cognitive structure provides strong anchorage for new informational units, whereas unstable and ambiguous
15
knowledge provide weak anchorage for new material wiiich will result ill poor learning.
In his discussion Ausubel distinguishes two independent dimensions of learning along which all learning can be located. The first dimension is the reception-discovery dimension. The essential feature of reception learning is Hiatt the content of the learning task is given to the student. Under this
circumstance the learner is required to comprehend the material and to
incorporate it into his cognitive structure. In discovery learning, on the other hand, the information to be learned is not given; the student must determine what inform ation is to be acquired before it can be established into the his/her cognitive structure. The importance of this first distinction becomes clear when we consider Ausubel’s claims that people as a whole acquire a massive body of knowledge primarily through reception learning.
The second contrasting dimension is the rote-meaningful continuum. The crucial factor in determining whether learning is rote or meaningful is the relevance or significance with which the new material is handled. Rote learning is a process in which new material is related to prior knowledge through arbitrary associations. That is, it involves the collection of new ideas bearing little or no relationship with the existing cognitive structure.
Meaningful learning is a process in which new material is related to prior knowledge through logical associations. Thai is, it involves meaningful interaction of new material with the existing cognitive structure. The very fact that material is associable with stable cognitive hooks accounts for its anchorage or meaningfulness.
The significance of the rote—meaningful learning continuum is perhaps best appreciated when we consider retention of learned material for long periods of time. Meaningfully learned material has far greater potential for retention, whereas rote learning is less productive and less retrievable. Try, for instance,
to recall where, who with, what you were doing on Feb 18, 1988. It is unlikely that you will remember anything, unless this date is relevantly associable with a special event in your life.
In short, the key concept in Ausubel's Assimilation Theory is that learning must be meaningful. The learners must understand what is to be learned; relate it to their present knowledge in a rational, nonverbaiini manner, and integrate fiie material being learned into their own cognitive structure.
I consider the learning process proposed by Ausubel clearly conscious. The assimilation of new knowledge is more feasible if it is relevantly associable with prior knowledge. It is the learner’s task to perceive relevant associations between what s/he already knows and what is to be learned. The learner, then, is in a position to deliberate what is learned or not. This decision involves consciousness.
To what extent does Ausubel’s cognitive theory apply to SL learning? Learning a second language involves both conscious and unconscious aspects, that will be unraveled throughout this dissertation. The conscious aspects are very well explained by the Assimilation Theory. However, the learning of an SL has peculiar aspects that develop in their own idiosyncratic way. Such aspects are known to undergo another type of learning, basically unconscious. This type is mostly explained by what I call linguistics-influenced theories, that will be discussed in the following chapters. But irrespective of the type of learning, Ausubel offers a powerful and thorough explanation of how knowledge builds up the cognitive structure.
17 REFERENCES
Ansubel, D., J. Novak & H. Hanesiatn. 1978. Educational Psychology: A Cognitive View. New York.Holt, Rinehart and Winston.
Bowers, K.S. & D. Meichenbaum (eds.) 1984. The Unconscious Reconsidered. New York:Wiley.
Cook, V. 1993. Linguistics and Second Language Acquisition.
London.M acM illan Press.
McLaughlin, B. 1978. The monitor model: some methodological considerations. Language Learning 28(2):309-32.
Reber, A.S. 1989. Implicit learning and tacit knowledge. Journal o f Experimental Psychology: General. 118(3):219-35.
Snow, C. In press. Bilingualism and Second Language Acquisition. In J. Berko Gleason & N. Bernstein Rainer eds.), psycholinguistics Today. New York: Holt, Rinehart and Winston.
THE MONITOR MODEL
la this chapter I briefly describe how the Monitor Model explains the second language acquisition process. After that, 1 develop the concept of acquisition process, or the internal cognitive ability responsible for the building up of language competence, as proposed by Krashen. Another aspect of the first hypothesis of the model, the acquisition-learning hypothesis, is discussed, then,the non-interface hypothesis. The final sections are concerned with criticisms of Krashen's model, as far as the first hypothesis is concerned. One criticism I raise is of the inconsistent use of the term subconscious. The others are related to Krashen's empirical evidence, or the lack of it.
The model
Krashen proposes a theory for building competence in a second language based on two independent processes, acquisition and learning. Acquisition is a subconscious process, similar or identical to the one children use to acquire the first and second languages in natural environments. It is sometimes referred to as an innate 'ability1 for the acquisition of languages. Learning is a / conscious process, involving internalization of the rules of the language and the ability to talk about them, normally brought about by instruction.
Besides this innate cognitive ability for the acquisition of languages, other conditions must be met for the acquisition process to take place, according to the Monitor Model: the adequate input (comprehensible input) must be provided to the learner, and the learner must be 'open' to this input by acquiring in an atmosphere in winch s/he is not anxious, thus keeping his/her 'affective filter* low. These conditions constitute the other hypotheses that form Hie Monitor Model, but they will not be developed here because they are not
19 relevant to tins study, since the focus here will be on the conscious/ unconscious issue.
The cognitive ability
According to Krashen et al. (1978:73), acquisition refers to the creative construction process. It is a product of the LAD — the Language Acquisition Device postulated by Chomsky, the difference being that the Chomskyan LAD did not depend on input to the same extent that Krashen's does: acquisition is a 'subconscious process for developing ability in language via die language "mental organ". [Chomsky 1975] Requires comprehensible input' (Krashen 1985:100). In this process, the learner already starts with basic assumptions or constraints on the language and develops along a pre-ordained sequence, known as the natural order. Krashen does not tell us much about this sequence, nor about Hie processes responsible for it.
One tiling that he does tell us about acquisition is that it involves hypothesis- testing, a process which takes place on a subconscious level of the mind (ibid:36). This process refers basically to the formulation of hypotheses, examining the input for confirmation or disconfirmation of the hypotheses (testing them), firming them up into rules, reorganizing the rules. Bley-Vroman (1986) and Snowr (in press) support the claim that the processes of hypothesis- testing and rale generation operate in the second language learner. Evidences listed by Snow (in press) are: (1) the occurrence of 'developmental errors' by second language speakers, i.e., errors which are explainable as overgeneralizations based on features of the target language rather than interference from features of the native language, for example, foots, hided, corned, (2) recapitulation of the same order of acquisition shown, by first language learners; (3) similar acquisition for speakers of different language
Despite the fact that Krashen does not explicitly address Universal Grammar (UG) when explaining his model, we cannot separate the workings of the LAD from the Chomskyan notion of UG. Krashen's reliance on Chomsky’s work, and the presence of a natural order suggest some kind of access to UG in Krashen's model. If so, then the hypothesis-testing ability is constrained by UG, that is, it can only produce a grammar consistent with the principles and parameters of UG. In oilier words, there is a built-in grammar that develops subconsciously when triggered by the adequate type of input, thus giving rise to a certain order of acquisition. For that reason, ungrammatical hypotheses (in reference to UG) are not feasible.
Researchers do not agree whether UG is available to the adult second language learner. There are several positions taken by them Some argue against tiie premise that UG is available to SL learners (Schachter 1990); some argue in favor of it (White 1990); some argue that there is evidence that SL learners have access to UG through their first language (Felix & Weigl 1991). In general, linguists' positions are similar to Krashen's: although they do not explicitly state it, some kind of access to UG is implied in their work.
One criticism made to the Monitor Model is related exactly to this point: because of more developed cognitive abilities, an adult can infringe upon the constraints of UG (Gregg 1984). In principle, the LAD corresponds to an initial stage of language acquisition in the child. In the adult, the LAD would not have
this characteristic. Concerning tins, Gregg (ibid SO) says:
Not only is an adult not in an initial state with respect to language, but he also is endowed with a much richer set of cognitive structures, which theoretically at least could enable him to violate the constraints of UG.
21
As a matter of fact, the acquisition-learning hypothesis ignores some characteristics which are particular to adults, such as superior cognitive capacities, larger memory span, meta-awareness of language, and more developed pragmatic knowledge, and assumes they are equal to children in this respect, stating that adults still have access to the LAD and so are able to use the acquisition process. This point is contested mainly by those who take a psychological view of the learning process. They assume that language knowledge is processed in the brain like any other kind of knowledge, that it is a product of general intellectual capacity, not specialized knowledge, and that as such it is influenced by the characteristics mentioned above as proper to adults
The non-interface hypothesis
A fundamental claim related to the acquisition—learning dichotomy is that learned knowledge cannot be transformed into acquired knowledge. This is known as the 'non interface hypothesis' (Krashen 1985:38). If acquisition is a process independent from learning, it does not benefit from conscious learning. In order to argue for this point, Krashen relies basically on observation of learners' reports. He cites examples of learners who are able to use complex structures in the second language and despite this, cannot explain the rules they use. The opposite is also true for him, that is, it is possible to find people who know the rides but are not able to use them correctly while actively using the SL. Finally, he also claims that SL learners or speakers use a lot more rules than they are able to describe, and this is true for very good learners or even for linguists. Such phenomena are part of the experience of every second language teacher, being, in tins sense, extremely intuitive. Actually, tins is what Krashen must have meant when he built lus theory: 'to explain phenomena and provide a sense o f understanding' (ibid: 104).
However, the statement learned competence does not become acquired competence* (ibid:42), also implies something that is not so empathetic to SL teachers, namely, that conscious knowledge cannot become subconscious. The process of internalization or routinization of structures, still common in SL classrooms, does not produce SL competence, according to Krashen On that aspect, he holds a very daring and provoking position, once so many teaching methods, approaches and techniques have relied so heavily upon these concepts for so long.
Aîi important consequence of the non-interface hypothesis is that learning serves only as a monitor, that is, an editor. Since it does not derive from the acquisition system, the monitor has a limited scope: it accounts only for those parts of the language for which the learner knows explicitly and consciously the rules. Besides, it is only available to the learner when s/he has time to monitor, and focuses on fomi, i.e., on the rules. Because the monitor works only as a source of editing or self-correction, it cannot start the production process (output). Only acquired knowledge serves for initiating utterances.
The conscious/subconscious issue
One of the most common criticisms made to the first hypothesis of the Monitor Model, the learning — acquisition hypothesis, is that Krashen has never defined the terms conscious and subconscious precisely, and it seems that this lack of precision in the definition has caused most of the problems in Krashen's proposal and consequently in the theory itself. In fact, these terms can encompass several definitions, depending on the context/area they refer to. In the case of Krashen’s model, if we consider the constructs that underlie the model, we will see that the concepts conscious and subconscious take us back to the Chomskyan notion of unconscious processes and knowledge, which are processes that are inaccessible in principle, that is, ’We can acquire knowledge
23 of their presence and activity by inference, but not by means of direct, immediate introspective awareness’ (Chomsky 1980 cited in Bowers and Meichenbanm 1984:156). Such processes and knowledge are very clearly defined by Chomsky: it is ’knowledge which resides in the mental system, and is actively employed in the service of ongoing cognitive processing, but which is incapable of being brought into phenomenal awareness (physical consciousness) and placed under voluntary control. We know the contents of the mind only by inference, never through direct introspection' (ibid). On that basis, I object to Krashen's use of the word subconscious, which, according to traditional definition in psychology (Reber 1985:740), refers to information that is at the margins of awareness, but which can be made conscious, given the proper circumstances. This type of information is not the same as tacit knowledge (see glossary), which is actually what Chomsky is talking about
t
(Chomsky 1975:164-6). Thus, it is wise to follow Reber as he adverts, subconscious 'should not, in any circumstance be used as a synonym for unconscious' (ibid).
Concerning the use of the terms unconscious/subconscious in tins dissertation, I followed the terminology used by each author in their models. Whenever my opinion was at matter, the term unconscious was preferred.
Empirical evidence
The most severe criticisms evoked by the Monitor Model are related to the lack of empirical data to support the leaming-acquisition distinction. McLaughlin (1978) states that the evidence for the two processes rests on the comparison between ’feel’ and ’knowing the rule’. It is true that Krashen compares acquisition to an intuitive feeling of grammaticality about the language, and learning to consciously evoking the corresponding rule. On this issue, McLaughlin (ibid:317) states that ’it is impossible to know whether
subjects are actually operating on the basis of ’rale’ or ’feel", above all because tliis is a subjective criterion In a response to his criticism, Krashen agrees that ’we have no physiological measure that shows an acquisition-learning difference’ (Krashen 1979:152) but he argues that like all research in Cognitive Psychology, iri which an abstract hypothesis is made and then checked to see if it accounts for real-world phenomena, 'the acquisition-learning distinction is an abstraction that predicts many observable and concrete phenomena’ (ibid). In this position, Krashen is supported by many important research leaders of the cognitive movement (Piatelli-Palmarini 1980).
Besides pointing out the lack of empirical evidence in Krashen's theory, McLaughlin (1978) also criticizes Krashen's evidence based on learner’s introspection. According to him, they are not reliable, since they cannot be tested empirically. In my view, this should not be pointed out as a weak point in Krashen’s theory, because Cognitivism makes room exactly for what Behaviorism refused to accept — the existence of a mental life and the introspective reports that account for it.
Hie most serious criticisms to introspective data in Krashen's model seem to be due to another cause, raised by Seliger (1983). When Krashen says that subconscious knowledge cannot become conscious, then this involves (again) the Chomskyan notions of competence and performance. The internal system of language or the competence system is unconscious, and as such, is unreachable. Our performance reflects this system, but is not necessarily identical to it, because other variables are involved in performance. When we try to describe tins internal system based on performance, we must be aware that we are describing an illustration of it, not the system itself. Because his theory is based on such theoretical tenets, Krashen could not, in principle, rely on introspective evidence to support lus theory.
25 Seliger (1983) refers to SLA research that relies on introspective reports as ’the psychoanalytic school of SLA’, in the sense that these retrospective reports on performance are to prove the learner's inner processes and his/her mental life. The main flaw of this line of research, according to Seliger, is that retrospective reports are taken as the product of competence or 'the acquired system', when, in fact, they are the product of performance. In other words, while they are elicited in order to show the process of language learning, in fact they show the product. This criticism makes sense, since competence in the Chomskyan sense (the sense being used here) is inaccessible through introspective awareness. For Seliger, once information is verbalized or brought to consciousness, it is already biased by interpretation and consequently different from what it was in the inner state.
In sum, Krashen proposes that two independent processes are involved in SLA: an subconscious one that leads to the formation of linguistic competence, based on internal innate capacities plus the provision of comprehensible input (or the focus of attention on communication rather than 011 linguistic forms); and a conscious one that serves as a monitor for self-correction, and that benefits from the formal teaching of rules. The fundamental claim of his model, known as the non-interface hypothesis, is that learning does not become acquisition, that is, the teaching of formal rules does not develop linguistic competence, it can only improve the monitor.
In the same way that Krashen’s theory seems intuitively right to many teachers and researchers, it is not considered a scientific theory by many others that claim it lacks empirical/experimental support. However, recent views of what can be considered ’scientific’ deny the need for ’positivist evidence’ as a quest for science. In the light of such views, Krashen’s theory would certainly be taken as an important contribution to the SLA field. This is specially true when we
consider his perceptions of the SLA process, which were forerunners of so many ideas, such as language specificity, unconscious knowledge, the non-interface hypothesis, relevance of meaningful input, and failure of the rote practice- based methodologies, that are .now widespread among SLA theories.
REFERENCES
Bley-Vroman, R 1986. Hypothesis testing, in second-language acquisition theory. Language Learning 36(3):353-75.
Bowers, K.S. & D. Meichenbaum (eds.) 1984. The Unconscious Reconsidered. New York:Wiley.
Chomsky, N. 1975. Reflections on Language. New York:Pantheon Books. Felix, S. & W. Weigi. 1991. Universal grammar in the classroom the effects of
formal instruction on SLA. Second Language Research 7(2): 162-81.
Gregg, K. 1984. Krashen's Monitor and Occam's Razor. Applied Linguistics 5(2):79-100.
Krashen, S. 1979. A response to McLaughlin, 'The monitor model, some methodological considerations’. Language Learning 29:151-67.
Krashen, S. 1985. The Input Hypothesis: Issues and Implications. New York: Longman.
Krashen, S., J. Butler, R Bknbaum & J. Robertson. 1978. Two studies in language acquisition and language learning. 1TL: Review o f Applied Linguistics 39-40:73-92.
McLaughlin, B. 1978. The monitor model, some methodological considerations. Language Learning 28(2):309-32.
Piatelli-Pahnarini, M. (ed.). 1980. Language and Learning: The Debate between Jean Piaget and Noam Chomsky. Cambridge, Massachusetts: Harvard University Press.
Reber, A 1985. The Penguin Dictionary o f Psychology. London:Penguin Books.
Schachter, J. 1990. On the issue of completeness in SLA Second Language Research 6(2):93-124.
Seliger, H. 1983. The language learner as linguist, of metaphors and realities. Applied Linguistics 4(3): 179-91.
Snow, C. In press. Bilingualism and second language acquisition. In J. Berko Gleason & N. Bernstein Ratner (eds ), Psycholinguistics Today. New York: Holt, Rinehart and Winston.
White, L. 1990. SLA and UG. Studies in Second Language Acquisition 12:121- 34.
CHAPTER 4
THE TASK-BASED APPROACH
Tins chapter deals with Prabhu's task-based approach. Actually, it is a pedagogical approach to the teaching of a second language rather than a second language theory or model. However, the assumptions about learning that underlie this pedagogical approach constitute a significant contribution to the present study of acquisitional processes. Firstly in this chapter, some of Prabhu’s assumptions about Hie acquisitional process, such as the system, the process, language awareness are presented. Such assumptions provide a better understanding of the classroom procedures, the second issue to be presented.
Thirdly, an analysis is made of the cognitive underpinnings of the task, followed by some concluding remarks.
The system
Prabhu (1987:69), proposes that learners have an internal system of rules which is activated whenever the learner is exposed to meaningful input in the second language, or in other words, when the focus of the learning task is on content, on meaning, on communication.
This internal system is not equal to the one described by generative grammarians, nor is it equal to any tentative description that linguists make of what they' consider to be the internal system: the internal system is much more complex than any descriptive grammar. It becomes clear in normal language use, when two levels of operation can be perceived: (1) a conscious one, where the language user conscious mind is busy with the message being communicated; and (2) an subconscious one, i.e., where his/her subconscious mind is simultaneously elaborating the message linguistically, in accordance ■with grammatical rules. The simultaneity7 of these two levels shows that the ability to elaborate the message linguistically does not develop directly and
consciously, but subconsciously, through focusing on meaning. Tliis is the basic assumption of learning and teaching that underlies the task-based approach.
Halliday (1985) perceives the unconscious nature of spoken language and its non-correspondence with construed grammars. He affirms that:
The sentence structure (of spoken language) is highly complex, reaching degrees of complexity that are rarely attained in writing. The categories of our language represent, unconscious slices of meaning; that is why it is so difficult to build a grammatical theory, because when people talk their unconscious choice of the language does not correspond to our conscious structuring of sentences and use of words. (pp.XXIV — XXVI; parenthesis added)
The task-based approach considers that the belief that the learner can acquire the internal svstem of a laneuaee bv consciously understanding and assimilating the grammar rules which have been described by linguists is equivocal. What Prabhu claims is that linguists’ grammars are conceptual, that is, they are not the language system per se but a probable picture of it, based on given outputs. In fact, outputs can provide just intuitions about the internal system; a picture of the system would be possible only if the system were isomorphic to available descriptions of it.
In this view of second language acquisition, 'planned progression' (Prabhu 1987:73), or the creation of a prospective syllabus based on language structures has no place, since language develops in spiie of and not because of the teaching of structures. It seems that this view is in accord with the notion that language follows an independent route of development when triggered by meaning-focused tasks. This route consists of transitional systems, as posed by Corder (1967) in his interlanguage studies. Corder's claim is that as systems, the different stages of interlanguage are holistic and not itemized, as prospective structural syllabi suggest.
29 The process
The task-based approach is based on the view that the second language system develops through an subconscious process, namely, siructure-abstraction. As defined by Prabhu (1987:69-70):
Task-based teaching operates with the concept that, while the conscious mind is working out some of the meaning-content, a subconscious part of the mind perceives, abstracts or acquires (or re-creates, as a cognitive structure) some of the linguistic structuring embodied in those entities, as a step in the development of an internal system of rules.
Besides the concept of structure-abstraction, deployment is another concept of fundamental importance ill Prabhu's view of language acquisition. It means that language abstracted while a task is being performed can be applied (used, transferred) to other contexts, being available for the purposes of production and understanding. Thus, it is said that linguistic competence achieved through the task-based approach is a deployable linguistic competence, because it is an ability that applies not only to immediate needs to express and understand meaning, but to the generation of language which is in accordance with grammatical norms.
In order for deployment in production to take place, it is believed, on Hie one hand, that abstract structures should be more firmly established than for comprehension. On the other hand, production serves to firm them up (ibid:?0-l).
Language awareness
The important point about language awareness within meaning focused teaching is that it is not enhanced or initiated externally by the teacher, but it arises spontaneously and naturally as part of the process of acquisition of a deployable internal system. Prabhu believes that this awareness somehow facilitates the learning task (ibid:76), but he does not explain in what way it does so.
According to the author, awareness as a natural consequence of the development of the internal system differs from externally-induced awareness in two aspects: it arises at its own time; and it arises at its own place, that is, awareness may show at a certain stage of the learning process which is intrinsic and subconsciously determined by the learner's internal system. As such, it is entirely unpredictable. It may be the case that it emerges in each learner at, a different time, and at a different stage of development, demonstrating individual variability. Thus, awareness induced by the teacher may not be very helpful if it does not overlap with the internal system's awareness of language. As posed by Prabhu (ibid:76):
Attention to form which is externally initiated or manipulated is likely to remain unrelated to either process (i.e., of meaning-extraction or of structure-abstraction) and can only be a pedagogic objective in itself. (parenthesis added)
The natural emergence of this awareness is undeniable, but we cannot say that it is determined by the internal system. As will be argued in the next chapters, metalinguistic awareness is a characteristic of the general cognitive maturity present in the adult or young adult. The fact that it emerges spontaneously while acquisition is taking place does not mean that it is a result of the linguistic competence system being formed. Albeit emerging spontaneously, it is still conscious, and thus cannot be the product of the structure-abstraction process.
The pedagogy of the task-based approach
A task is defined as ’an activity w'hich required learners to arrive at an outcome from given information through some process of thought and which allowed teachers to control and regulate that process' (ibid:24). According to the task- based approach, class activities should offer a cognitive challenge, that is, they should use cognitive processes such as reasoning, inferring, and inter-relating
31 information in a logical way, so as to provide the learner with a procedure in which s/he could focus entirely on meaning.
The class in the task-based pedagogy7 consists basically of three stages: the pre-task, the task, and the marking of the student's outcome of the task by the teacher. The pre-task is a whole-class activity guided by the teacher, which consists of a task that is solved by the teacher together with the class, in the form of interaction and negotiation. The teacher uses questions or instructions on the task in order to explain it. A task similar to this one is then given to the students in the ’task' part, for each one to solve it individually and in writing the only interaction being between the learner and the task. In the third part of the class, the teacher gives feedback to the students, by marking, their work not on language form, but on content. Tins part serves also as a guidance to the teacher in terms of Hie level of challenge that Hie task represents.
The cognitive counterpart
Prabhu compares subconscious abstraction of linguistic structures with the formation of cognitive structures. Cognitive structures can be inaccurate and incomplete at first, but with constant exposure to the target language, with constant effort to extract meaning from the same piece of language, which appears repeatedly during the three pedagogic stages, they are modified, thus becoming accurate, complete and firm.
Not only Hie abstraction of structures is possible through meaning-focused activities, but their 'elaboration' also benefits from it. Existing cognitive structures make it possible for new linguistic structures from new language samples to be understood and in this process the existing structures suffer an adjustment, that is, they may be firmed up, modified, extended, or integrated. In this way the internal system is developed, little by little.
This view of language acquisition resembles some notions developed by Ausubel (1978) in his view of acquisition of knowledge, the Assimilation
Theory. In fact, a parallel can be drawn between the theories. Like Prabhu, Ausubel establishes that there is a cognitive structure and there follows hierarchically anchoring ideas, which are part of the cognitive structure. The most important requisite to anchor a new idea is the availability' of previous knowledge (anchoring ideas). Anchoring ideas also suffer a process of adjustment as they are exposed to new data They may be enlarged, changed or integrated to new ideas. The notion of deployment can also be compared to a similar notion in Ausubei’s theory, namely, the notion of transferability of newly learned material to the existing cognitive structure. Transfer, as defined by Ausubel (1978:166) refers to 'the impact of prior experience upon current learning'.
In spite of such similarities, Prabhu's concept of learning differs intrinsically from Ausubel's. Ausubel maintains that the cognitive structure is constructed, not activated, via interaction with the environment, in a very conscious manner, that may include practice as well. It relies principally on the environment rattier than on the internal system. In fact, for Ausubel, there is no internal system to be triggered, whai there is is an ability to acquire the language system. As we will see, tins constitutes a fundamental difference between theories/models that rely on linguistic assumptions and those which rely on psychological ones.
The cognitive underpinnings of the task: problem-solving
The task-based approach can be said to use a cognilive-oriented procedure, the task. It is so because the task involves a cognitive device known as problem
solving. Problem-solving is defined by Ausubel as 'any activity in 'which both the cognitive representation of prior experience and the components of a current problem situation are reorganized in order to achieve a designated objective' (ibid:565). It involves sequentially two kinds of learning: reception and discover}7 learning.
